historical development Jonathan Geary, University of Arizona - - PDF document

7/13/2019 1

The Piipaash word for ‘coyote’ as a window into Yuman historical development

Jonathan Geary, University of Arizona jonathangeary@email.arizona.edu http://u.arizona.edu/~jonathangeary/ SSILA Summer Meeting 2019 July 13, 2019

Special thanks to………

• The Piipaash Elders and the broader Piipaash community for having

shared their language with me.

• Skye Anderson
• Luis Barragan
• Amy Fountain
• Max Mulé
• John Powell
• Kelly Washington

2

1 2

7/13/2019 2

Introduction

• I hypothesize that the Piipaash word xatlywe /xatʎwe/ ‘coyote’ derives

from historical reflexes of two independent morphemes: xat /xat/ < /*xat/ ‘dog’ and lyvii- /ʎviː-/ < /*ʎwi(ː)-/ ‘be like, look like’.

• That is, xatlywe more literally refers to something which ‘looks like a dog’.
• This is supported by the use of xatlywe in archival materials, wherein Piipaash

speakers extend it to other Canidae, namely ‘fox’ and ‘wolf’.

• I show that other Yuman languages similarly derive a common name for

‘coyote’ from that for ‘dog’, which is distinct from the name of mythic Coyote.

• The /v/ of lyvii- reflects a shift from Proto-Yuman /*w/ > River /v/. Xatlywe

became lexically frozen prior to this shift, hence the retention of /w/.

• Xatlywe provides evidence for the direction of the proposed shift, which both

is rare cross-linguistically and has had important historical implications.

3

Piipaash

• Piipaash (mrc; 33.51, -111.75) (also spelled “Pee Posh”; a.k.a. Maricopa) is a

Yuman language spoken near Phoenix, AZ (Ethnologue s.v. “Maricopa”).

• Yuman languages are today spoken in Arizona, California, and Baja California.
• Together with Mojave (mov; 34.89, -114.6) and Quechan (yum; 32.79, -114.6),

Piipaash forms the River subbranch of the Yuman family (e.g. Miller 2018).

4

Figure 1. Yuman language family tree (based on Miller 2018).

3 4

7/13/2019 3

Piipaash

• Piipaash (mrc; 33.51, -111.75) (also spelled “Pee Posh”; a.k.a. Maricopa) is a

Yuman language spoken near Phoenix, AZ (Ethnologue s.v. “Maricopa”).

• Yuman languages are today spoken in Arizona, California, and Baja California.
• Together with Mojave (mov; 34.89, -114.6) and Quechan (yum; 32.79, -114.6),

Piipaash forms the River subbranch of the Yuman family (e.g. Miller 2018).

• The U.S. Census Bureau estimated that there were 35 speakers of

“Cocomaricopa” in 2015. However, more recent estimates place the number of speakers much lower (e.g. SRPMIC Cultural Resources Department, n.d.).

• Community members are fluent in English, some also in Akimel O’odham (Uto-Aztecan).
• Revitalization efforts are ongoing at both the Salt River Pima-Maricopa Indian Community’s

O’odham Piipaash Language Program and the Gila River Indian Community.

5

Observations from archival materials

• The names for flora and fauna are occasionally glossed inconsistently in

archival materials: they may be inconsistent across speakers/fieldnotes, or inconsistent with how today’s speakers use the same words.

• Some of these inconsistencies reflect misidentifications, most likely on the

part of the linguist, who may be unfamiliar with the local wildlife and/or unable to make the relevant connections (e.g. Rea 2007: xvi-xvii).1

• Whereas Langdon et al. (1991) and others have recorded xnvchiip as ‘small sparrow’,

Spier (1946: 113) glosses the same word as ‘cactus wren (?)’.

• James Crawford (1962) worked with one speaker who identified chyer ‘bird’ as ‘bluejay’.
• But some of these inconsistencies reflect legitimate differences in how

speakers of the past used these words, and these differences can provide valuable insights into how Piipaash has changed over time………

6

Cactus wren House sparrow

1 In respect of Piipaash customs which prohibit using the names of the deceased, I refer to the linguists who collected these notes instead.

5 6

7/13/2019 4

The ‘coyote’ problem

• In modern Piipaash, xatlywe /xatʎwe/ is used only as the common

name for ‘coyote’ (Langdon et al. 1991). However, archival materials attest to this name being extended to ‘fox’ and ‘wolf’ as well.

• In field notes dating from 1929 to the early 1960s (collected by James Crawford,

Judy Crawford, Alfred Kroeber, and Leslie Spier) xatlywe is glossed as ‘fox’ either in

addition to or in place of ‘coyote’ a number of times.

• Additionally, at least two speakers are attested as having used xatlywe as a

generic term for ‘coyote, fox, and wolf’ (Spier 1946: 105; Alpher 1970).

• cf. modern Piipaash qoqo/mkwe ‘fox’, xatkuuly ‘wolf’ (Langdon et al. 1991).
• This latter grouping includes all of the major Canidae which Piipaash speakers

would have been familiar with, with the exception of xat ‘dog’.

7 Wikimedia Commons

The ‘coyote’ problem

• Why are speakers attested as having used xatlywe historically for ‘fox’

and ‘wolf’, in addition to ‘coyote’? There are two possibilities:

• 1. These are mistakes. Someone involved in recording these materials must

have misidentified xatlywe as ‘fox, wolf’.

• Spier thought so: Near the end of his notes (1929-1930), he wrote that earlier

glosses of xatlywe needed to be corrected from ‘fox’ to ‘coyote’.

• 2. These are valid uses of xatlywe. Historically it served as a generic term for

non-dog Canidae, with its meaning narrowing to ‘coyote’ over time.

8 Wikimedia Commons

7 8

7/13/2019 5

The ‘coyote’ problem

• Why are speakers attested as having used xatlywe historically for ‘fox’

and ‘wolf’, in addition to ‘coyote’? There are two possibilities:

• 1. These are mistakes. Someone involved in recording these materials must

have misidentified xatlywe as ‘fox, wolf’.

• Spier thought so: Near the end of his notes (1929-1930), he wrote that earlier

glosses of xatlywe needed to be corrected from ‘fox’ to ‘coyote’.

• BUT! Coyote is an important character in Piipaash folklore (e.g. Spier 1933).
• AND! Coyotes are not uncommon in the Southwest. It would be strange for

Piipaash speakers or even American linguists to misidentify them………

• 2. These are valid uses of xatlywe. Historically it served as a generic term for

non-dog Canidae, with its meaning narrowing to ‘coyote’ over time.

9 Wikimedia Commons

Two observations about xatlywe

• 1. The mythic character Coyote has a different name which is unrelated

to the common name xatlywe (Spier 1933: 353; Wares 1968: 81; Alpher 1970).

• This is true in other Yuman communities, such as Yavapai (Kendall 1980: 132).
• Traditional Piipaash customs prohibit using the names of living and (especially)

deceased individuals (Spier 1933: 197-198) (hence I avoid using the mythic name here).

• 2. The first syllable of xatlywe overlaps with the morpheme xat ‘dog’.
• This is also true of the common name xatkuuly ‘wolf’.
• It’s tempting to hypothesize that both xatlywe ‘coyote’ and xatkuuly ‘wolf’ are

derived from xat ‘dog’. This might even explain why xatlywe is never extended to ‘dog’ in archival materials………

10

9 10

7/13/2019 6

Hypothesis

• I hypothesize that the common name xatlywe ‘coyote’ derives from

historical reflexes of xat ‘dog’ and the verb lyvii- ‘be like, look like’.

• That is, xatlywe more literally refers to something which ‘looks like a dog’.
• Speakers combined these morphemes to form a new word which could have

plausibly referred to any non-dog Canidae, and which they used to refer to the common ‘coyote’ in order to avoid using the mythic name.

• The historical use of xatlywe for ‘fox, wolf’ reflects speakers’ sensitivity to the

morphological composition of this word (these animals too look like a dog).

• Over time, xatlywe became lexically frozen and narrowed in its scope, such

that speakers now can only use it to mean ‘coyote’.

11

Evidence for xat ‘dog’ as a component of xatlywe

• Dogs (xaat) were an important pet kept by the Piipaash (Spier 1933: 73-74).
• One of Spier’s consultants attested that dogs were in several senses human

(hence their meat could not be eaten, unlike other animals kept as pets), and they could even appear in dreams in the form of humans (Spier 1933: 74; 254).

• Did Piipaash speakers use xat to derive other names? If so, it would make it

more plausible that they also used xat to create xatlywe.

• nyxat ‘pet’ (ny-xat ‘POSS-dog’) (Spier 1933: 73; Langdon 1978; Langdon et al. 1991).
• xatkuuly ‘wolf’ (Spier 1946; Langdon et al. 1991) – Spier (1946: 106) hypothesizes

that it means ‘bigger than a dog’, but I suspect that the second morpheme is kuly- ‘climb, go up’, making xatkuuly literally a ‘dog that climbs’.

• txpa(sh) ‘Akimel O’odham/Pima’ (Spier 1933: 7; Langdon et al. 1991) – I suspect

that this is metathesized from *xtpa(sh) (xt-pa-sh ‘dog-person-PL’).

• txpa xat ‘dog Pima’, another O’odham-speaking group (Spier 1933: 7).

12

• 11

12

7/13/2019 7

Evidence for xat ‘dog’ as a component of xatlywe

• In fact, Piipaash seems to have followed a Yuman-wide strategy of deriving

a common name for ‘coyote’ using the morpheme for ‘dog’:

• I surveyed other Yuman languages + Cochimí (which Mixco (1978) identifies as

a distant relative of Yuman) for the words for ‘dog’ and ‘coyote’.

• In 3/4 subfamilies + Cochimí, at least one language appears to derive the

word for ‘coyote’ from that for ‘dog’ plus another morpheme.

13

Table 1. Yuman words for ‘dog’ and ‘coyote’.

River Pai Delta-California Kiliwa2 Cochimí7 Piipaash1 Quechan2 Mojave2,3 Yavapai4 Cocopa5 Ipai6 ‘dog’ xat 'axáṭ (‘horse’) 'ahat (‘pet’), hatchoq kθár xaṭ 'ehatt thát Ɂ-t-(k)at ‘coyote’ xatlywe xatalwé hukthar kθàrʔhāna xt̤pa hattepaa miltiˀ Ɂ-t-(x)at-wachitaba-wa

The data presented in this table comes from the following: 1 Langdon et al. (1991); 2 Wares (1968); 3 Munro et al. (1992); 4 Shaterian (1983);

5 Crawford (1989); 6 Couro and Hutcheson (1974); 7 Mixco (1978). As I aim only to demonstrate similarities across forms within individual

languages, I have not attempted to regularize the orthographies used here.

Evidence for xat ‘dog’ as a component of xatlywe

• Given that speakers have formed other words from xat ‘dog’, it is plausible

that they also derived xatlywe ‘coyote’ this way.

• Given (1) that a similar form exists in Quechan xatalwé ‘coyote’, (2) that

these languages are closely related, and (3) that historical animosities had existed between the two communities (Kroeber and Fontana 1986) which likely precluded contact/borrowing between them, it seems plausible that xatlywe dates back at least to speakers of Proto-Maricopa-Quechan.

• It could also be that xatlywe dates back to Proto-River, with Mojave hukthar

having replaced it as a form borrowed from the neighboring Pai languages (cf. Hualapai gathád [gaθaɾ] (Watahomigie et al. 2003); Yavapai k𝜄ár ‘dog’ (Shaterian

1983) (see also Wares 1968: 81)), but this is entirely speculative.

• For simplicity, I will continue to treat this as a Piipaash phenomenon.

14

13 14

7/13/2019 8

Phonological developments in River Yuman

• So if Piipaash xatlywe ‘coyote’ is related to xat /xat/ ‘dog’ and lyvii-

/ʎviː-/ ‘be like, look like’, then why is the word not *xatlyvii?

• I propose that xatlywe involved historical forms /*xat/ and /*ʎwi(ː)/.
• /*xat/ is a fairly uncontroversial reconstruction of the Proto-River form for

‘dog’ (cf. Mojave ahat ‘dog, pet’, Quechan 'axáṭ ‘dog, horse’ (Wares 1968)), which has remained unchanged in Piipaash xat and xatlywe.

• /*ʎwi(ː)/ must have undergone two changes, both of which have been

proposed for the River subbranch previously:

• Lowering of /*i(ː)/ to /e/ in xatlywe (Langdon 1976; Munro and Gordon 1990);
• Strengthening of /*w/ to /v/ in lyvii-, but not in xatlywe (Langdon 1975).

15

Lowering of high vowels in River Yuman

• Langdon (1976) proposed a three-vowel system for Proto-River: /*i, *u, *a/.
• Furthermore, she proposed that the high vowels /*i, *u/ lowered to /e, o/

except before palatal consonants or /k, s/ (see also Munro and Gordon 1990).2

• The /*i/ of /*ʎwi(ː)/ would have lowered to /e/ in xatlywe. However, the

vowel remained /iː/ in lyvii-, which often cooccurs with the -k realis suffix.

• Similar /e~iː/ alternations can be found in Piipaash (and in Mojave, Quechan):
• 1. we-xa ‘do-IRR’

wii-k ‘do-say.COMP’

(Munro and Gordon 1990)

• 2. kwe ‘cloud’

kwii- ‘be cloudy’

(Langdon et al. 1991)

• 3. xwe ‘enemy, war’ xwii- ‘be enemies, have war’ (Langdon et al. 1991)

16

2 Working with limited data beyond Quechan, Langdon (1976) originally framed this as a process which had applied in Quechan. However,

this rule can and has also been applied in explaining phonological developments in Mojave and Piipaash (Munro and Gordon 1990).

15 16

7/13/2019 9

Strengthening of /*w/ to /v/ in River Yuman

• A major sound correspondence differentiating River Yuman from the other

subbranches is that between /w, j/ and River /v, ð/ (Kroeber 1943, Wares 1968).

• Langdon (1975) proposed that these sounds correspond to Proto-Yuman

/*w, *j/, which strengthened to /v, ð/ root-initially in River Yuman.

17

Table 2. Correspondence sets for non-River /w, j/ vs. River /v, ð/ (data from Wares 1968). Kiliwa Pai Delta-California River Word Yavapai Cocopa Piipaash Mojave Quechan ‘house’ wáˀ uwá wá aváˀ ˀavá ˀavá ‘stone’ wé·y ˀuwíˀ wí ví· ˀaví· ˀaví· ‘tooth’ yá·w ˀyó·ˀ iyá iđó i·đó· i·đó·

I have maintained Wares’ original transcriptions. The letter ‹y› corresponds to /j/, and ‹đ› corresponds to /ð/.

Strengthening of /*w/ to /v/ in River Yuman

• I propose that xatlywe became lexically frozen prior to the /*w/ > /v/ shift

in River, preserving the original consonant of /*ʎwi(ː)/ (> Piipaash lyvii-).

• Phonological shifts whereby glides strengthen to fricatives (rather than the

reverse) are rare cross-linguistically (Pulleyblank 1989: 388; Shaul and Hill 1998: 380). In fact, the rarity of this change has been used as evidence:

• for grouping Mojave, Quechan, and Piipaash into a sub-family of Yuman.
• that speakers of Proto-River and Proto-Tepiman, which exhibits a similar shift

/*w, *j/ > /g, d/, were in contact in the Hohokam period: glide strengthening diffused from one speech community to the other (Shaul and Hill 1998).

• The preservation of /*w/ in xatlywe substantiates this shift in River Yuman,

evidence for the direction of which has until now been lacking.

18

17 18

7/13/2019 10

Loose ends

• Have Piipaash speakers ever used lyvii-/-lywe to form other words?
• To my knowledge, no. There are no other words in Langdon et al. (1991) which

bear -lywe, nor am I aware of other such words in Piipaash.

• This is not unusual: plenty of other animal names decompose into at least
• ne morpheme which doesn’t recur in others (e.g. xatkuuly ‘wolf’).
• Could Piipaash speakers have borrowed xatlywe?
• We have seen that other Yuman languages, except Quechan, use a different

word for ‘coyote’. Even if Piipaash speakers borrowed xatlywe from Quechan (or vice versa), it still originated within the Maricopa-Quechan sub-family.

• The Piipaash and O’odham (Uto-Aztecan) have long lived together (Spier 1933),

but the relevant names in O’odham are different (cf. ban ‘coyote’, gogs ‘dog’

(Zepeda 1984)), so it is unlikely that they influenced xatlywe.

19

Summary

• In archival materials, Piipaash speakers are attested as having extended xatlywe

‘coyote’ to ‘fox, wolf’ as a generic term for non-dog Canidae.

• I have proposed that xatlywe is derived from historical reflexes of xat ‘dog’ and

lyvii- ‘be like, look like’: namely, /*xat/ and /*ʎwi(ː)/.

• Piipaash has followed a Yuman-wide strategy of deriving a common name for

‘coyote’ from that for ‘dog’, possibly to avoid using Coyote’s mythic name.

• While xat remained the same, /*ʎwi(ː)/ underwent (1) lowering of the vowel

to /e/ in xatlywe, and (2) strengthening of /*w/ to /v/ root-initially in lyvii-.

• /*w/ > /v/ is a characteristic feature of River Yuman, and the direction of this

shift has significant historical implications (Shaul and Hill 1998). Xatlywe became lexically frozen prior to this shift, and so preserves the /w/. This etymology of xatlywe thus provides crucial evidence for the direction of this shift.

20

19 20

THE PIIPAASH WORD FOR ‘COYOTE’ Geary 1

The Piipaash word for ‘coyote’ as a window into Yuman historical development Jonathan Geary (jonathangeary@email.arizona.edu), University of Arizona References Alpher, Barry. 1970, May 5. Unpublished fieldnotes. University of California, Berkeley: The Bancroft Library. Couro, Ted, and Christina Hutcheson. 1974. Dictionary of Mesa Grande Diegueño. Malki Museum Press. Crawford, James M. 1962. Maricopa word lists (unpublished fieldnotes). American Philosophical Society. Crawford, James M. 1989. Cocopa dictionary. Berkeley, CA: University of California Press. Kendall, Martha B. 1980. Coyote marries his daughter (Yavpe). In Martha B. Kendall (ed.), International Journal of American Linguistics Native American Texts Series: Coyote Stories II, p.129-133. Ann Arbor, MI: Imprint Series, UMI Monographs. Kroeber, Alfred L. 1943. Classification of the Yuman languages. University of California Publications in Linguistics 1: 21-40. Kroeber, Clifton B., and Bernard L. Fontana. 1986 (reprinted 1992). Massacre on the Gila: an account of the last major battle between American Indians, with reflections on the origin of

• war. Tucson, AZ: The University of Arizona Press.

Langdon, Margaret. 1975. Boundaries and lenition in Yuman languages. International Journal of American Linguistics 41: 218-233. Langdon, Margaret. 1976. The Proto-Yuman vowel system. In Margaret Langdon and Shirley Silver (eds.), Hokan Studies: Papers from the First Conference on Hokan Languages, p.129-

• 148. The Hague: Mouton.

Langdon, Margaret. 1978. The origin of possession markers in Yuman. In James E. Redden (ed.), Proceedings of the 1977 Hokan-Yuman Languages Workshop, p.33-42. Langdon, Margaret, et al. 1991. Maricopa – English dictionary. Prepared by Margaret Langdon and others. SRPMIC Department of Education. Miller, Amy. 2018. Phonological developments in Delta-California Yuman. International Journal

• f American Linguistics 84: 383-433.

Mixco, Mauricio J. 1978. Cochimi and Proto-Yuman: lexical and syntactic evidence for a new language family in Lower California. Salt Lake City, UT: University of Utah Press. Munro, Pamela, Nellie Brown, and Judith G. Crawford. 1992. A Mojave dictionary. UCLA Occasional Papers in Linguistics 10. Munro, Pamela, and Lynn Gordon. 1990. Inflectional ablaut in the River languages. In Scott DeLancey (ed.), Papers from the 1989 Hokan-Penutian Languages Workshop, p.69-86. University of Oregon Papers in Linguistics. Pulleyblank, Edwin G. 1989. The role of coronal in articulator based features. In Caroline Wiltshire, Randolph Graczyk, and Bradley Music (eds.), Papers from the 25th Annual Regional Meeting of the Chicago Linguistic Society, p. 379-393. Rea, Amadeo M. 2007. Wings in the desert: a folk ornithology of the Northern Pimans. Tucson, AZ: The University of Arizona Press. Shaul, David Leedom, and Jane H. Hill. 1998. Tepimans, Yumans, and other Hohokam. American Antiquity 63: 375-396. Shaterian, Alan. 1983. Yavapai phonology and dictionary. PhD dissertation, University of California, Berkeley.

THE PIIPAASH WORD FOR ‘COYOTE’ Geary 2

Spier, Leslie. 1929-1930. Unpublished fieldnotes. Manuscript Collection, Museum of Northern

• Arizona. MS-118-2-1-4, MS-118-2-1-5, and MS-118-2-1-6.

Spier, Leslie. 1933 (reprinted 1978). Yuman tribes of the Gila River. New York, NY: Dover. Spier, Leslie. 1946. Comparative vocabularies and parallel texts in two Yuman languages of

• Arizona. Albuquerque, NM: The University of New Mexico Press.

Wares, Alan Campbell. 1968. A comparative study of Yuman consonantism. The Hague: Mouton. Watahomigie, Lucille J., Jorigine Bender, Malinda Powskey, Josie Steele, Philbert Watahomigie Sr., and Akira Y. Yamamoto. 2003. A dictionary of the Hualapai language. Zepeda, Ofelia. 1984. Topics in Papago morphology. PhD dissertation, University of Arizona.